This is an over view of response times for Eugene Police Department (EPD). Source at Github
Dispatch data was retrieved from Eugene Police Department’s online dispatch log portal:
Cartographic elements are from “Oregon geospatial data clearinghouse” https://www.oregon.gov/geo/Pages/sdlibrary.aspx <ftp.gis.oregon.gov>
all calls
The maximum amount of calls the portal will return is 250.
Since the script that scraped the search portal querried each day seperatly, and looking at the number of logs retrieved for each day in the above chart, we can see that we didn’t hit that 250 log limit so our data set isn’t missing logs on busy days.
The FBI categorizes the a handful of crimes as Part 1.
“The UCR Program collects data about Part I offenses in order to measure the level and scope of crime occurring throughout the nation. The program’s founders chose these offenses because they are serious crimes, they occur with regularity in all areas of the country, and they are likely to be reported to police.” -FBI Definitions
Part 1 offenses are
These names don’t directly correspond to the incident descriptions in the EPD dispatch data set. Instead they map to:
noting that homicides and rape are not present in the EPD dispatch dataset.
overview of total number of calls for serious crimes.
Since we are missing catagories, let’s see what the Federal NIBRS reports look like.
| NIBRS Crime Description | totals |
|---|---|
| Aggravated Assault | 1105 |
| Arson | 342 |
| Burglary/Breaking And Entering | 2808 |
| Forcible Rape | 238 |
| Motor Vehicle Theft | 2225 |
| Murder and Non-Negligent Manslaughter | 6 |
| Robbery | 444 |
| Theft From Building | 732 |
| Theft From Motor Vehicles | 6260 |
| Incident Desc | calls |
|---|---|
| Arson | 16 |
| Assault 4 | 543 |
| Burglary | 6030 |
| Robbery | 550 |
| Theft | 4972 |
Both the dispatch calls and federal reported incidents are close to
being normally distributed, but there are some outliers so I looked at
both the Person and Spearman correlation coefficients, which are 0.64
and 0.65 respectively indicating a moderately strong correlation. Thus
despite the Dispatch Calls not fully covering the extent of all reported
crimes, we can still use those calls since they are strongly correlated
with the federal data.
##
## Attaching package: 'zoo'
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric
| x | |
|---|---|
| Estimate | 0.1172015 |
| Std. Error | 0.0512320 |
| t value | 2.2876614 |
| Pr(>|t|) | 0.0258220 |
## `geom_smooth()` using formula 'y ~ x'
## Warning: Removed 6 row(s) containing missing values (geom_path).
| x | |
|---|---|
| Estimate | 0.0038033 |
| Std. Error | 0.0032665 |
| t value | 1.1643290 |
| Pr(>|t|) | 0.2490577 |
## `geom_smooth()` using formula 'y ~ x'
## Warning: Removed 6 row(s) containing missing values (geom_path).
## 5 year trendlines, Part 1 calls
## `geom_smooth()` using formula 'y ~ x'
## `geom_smooth()` using formula 'y ~ x'
## Warning: Removed 6 row(s) containing missing values (geom_path).
Calls per day and average response times by month for 5 year(2018 - 2022), 3 year () and 1 year (2022) time windows. below.
Next is an overview of the highest, lowest and mean for each month for
average daily call response times for part 1 crimes.
| Incident Desc | total_calls | avg_resp |
|---|---|---|
| Theft | 4972 | 145 |
| Burglary | 6030 | 87 |
| Robbery | 550 | 36 |
| Arson | 16 | 35 |
| Assault 4 | 543 | 32 |
| Incident Desc | total_calls | avg_resp |
|---|---|---|
| Theft | 690 | 149 |
| Burglary | 884 | 68 |
| Arson | 3 | 29 |
| Robbery | 48 | 25 |
| Assault 4 | 79 | 22 |
All Pt1 Calls